In order to perform these simulations, there is a need for IT and computational biologists. The biologists and the medical doctors have to very closely talk to each other. The centre now has about 250 people. Half of them have a background in experimental biology and the other half in computational biology, IT, physics, and so on. These people are working in two buildings. Recently, part of them moved into a second building. Ninety percent of the researchers are in one building, spread out over three floors, but it is key that they are in one building at the new campus of Belval where most of the University of Luxembourg is located since a couple of months.
As for the computational resources that are available, Rudi Balling told that they have about 1000 cores and that the centre is storing 5 Petabytes of data. The amount of data that is coming out of analyzing biological systems as well as medical disease systems such as imaging, DNA sequencing, analyzing with robots, is just exploding. Within two years, the researchers at Belval have about a tenfold of data more than they had a few years ago.
If you look into the IT infrastructure in Luxembourg it is pretty good. Luxembourg has a finance industry that needs to have quick online retrieval of information from the financial markets and the stock market. Luxembourg has very good networking connections to Paris and London. The country has excellent data storage centres. There is infrastructure in IT, already available, that needs to be continuously built, but it can be capitalized and built on, so Luxembourg doesn't need to build anything from scratch. The value of infrastructure - initially, it was more in the area of computational biology and sequencing centres - is now the key issue if you want to compute.
International collaboration is essential, particularly when you are in a small country like Luxembourg. This country has about half a million people, five hospitals, and is 80 by 60 kilometers long. Luxembourg cannot compete, for example, by studying large patient numbers such as France and the United Kingdom. Rudi Balling explained that the researchers have to bring in additional elements. Maybe they can analyze the data better than in other countries. For researchers in Luxembourg information and communication technology, modelling, simulation, network analysis, disease analysis are areas where they focus on and where they would like to put a stamp on in the international competition.
Rudi Balling thinks the e-Infrastructure Commons is a fantastic idea, something that the community really has to develop. He considers infrastructure, especially in IT, is an attractor. It has done more for the European Research Area (ERA) than anything else. It brings people together, you don't have to reinvent the wheel, you can talk to experts. Infrastructure is the place where people need to discuss what to do with the infrastructure, almost like a form, and if the infrastructure is at the edge of technology, it has a very catalytic effect. You need to have very interesting pilot projects brought to the front end of the infrastructure and then, things will move.
One of the challenges Rudi Balling sees in the ICT revolution, the avalanche that is coming in with Big Data, complex data, heterogeneous data, is that the infrastructure needs to sort of compose the effect by algorithms that help us to analyze the data. Fast computing is becoming more and more essential but fast computing will be driven by two major factors. One is our ability to have fast computers. The other is our ability to develop algorithms that take advantage of these fast computers. There is a need to invest in both: software developers who are able to bring the projects to these high performance computing facilities and centres. The users, the experimentalists, the application developers have to be very close to the people who develop the latest infrastructure. What kind of computers do we need, because more and more the algorithms are built-in into the computer. There needs to be a very close interaction between experimentalists and computer experts.
When you see what is now happening in the area of machine learning and machine intelligence, a major incentive of this is how the human brain works. Lots of information is coming in, lots of it is not completely reliable. You need to have a lot of information that is integrated, you need to weigh the liability of the information, you need feedback from what you have seen in the past, you need to project it with a probability into the future. Computer programmes in the future will partly learn from the human brain and vice versa.